Kyoto Culture Forum / QUESTION 01
AI makes meeting easier. What about the relationship?
Translating, preparing an exchange, or finding someone to talk to has become more accessible. Trust, reciprocity, and the knowledge that comes from meeting someone are still built between people.

When AI makes a professional meeting easier, does it change the form of networking or the nature of the relationship?
Abdel Noubli's response
How AI makes a first meeting easier says nothing about the quality of the relationship that follows.
Access to the encounter
For someone starting out, who doesn't speak the language, or who knows little about the field, AI can help prepare for a meeting, phrase an approach, and grasp the context's unwritten rules.
Trust takes shape in the exchange
A convincing pitch can be generated. Lived experience, a third party's word, and the ability to respond to what's happening in the conversation count differently: they give reasons to trust.
The criterion proposed to the group
To judge a relational tool, look at what it brings about. Does it make possible a human meeting that would otherwise have been difficult, or does the interaction end within the tool itself?
Networking — Current Situation Analysis
PREAMBLE
I was given the Networking domain, I believe, because of the ground on which I have had to practise it: twenty years as a serial entrepreneur, companies built in France, the United Arab Emirates and the United States, and work carried out across Europe, North Africa, and — most recently — Japan and Taiwan.
What strikes me, having done this on three continents, is how similar networking is everywhere. The forms differ profoundly: the exchange of meishi with two hands in Tokyo is not the handshake of a Paris café, nor the elevator pitch of a San Francisco conference. But the function is the same.
And that function is not only business. Networking is also belonging, mutual support, learning, and reputation — it is how a person stops being alone. Yet it carries an economic consequence we should not hide: it is one of the main mechanisms by which opportunity is distributed. Whoever has a network has access; whoever does not, does not.
What every culture I have worked in seems to know — and Japan perhaps most explicitly — is the order of things: the relationship comes first, the business follows. Trust is built before it is used. That ordering is precisely what is at stake when AI enters the encounter.
So the question I would put to this group is this:
If the function of networking is universal but its forms are deeply cultural, what exactly does AI change — the form, or the function?
My answer, in seven points.
01 Current state: maturity is uneven, and the gap is widening
Already mature and largely invisible: AI-assisted outreach and messaging, real-time translation, matchmaking algorithms on professional platforms, CRM enrichment, automated meeting summaries. These are now embedded in how professional relationships begin — often without either party naming it.
Still entirely immature: everything that constitutes the substance of a relationship — trust formation, reciprocity, obligation, presence, the unplanned encounter. No system produces these. They remain wholly human.
The asymmetry matters: AI has industrialized the entry into relationships while leaving their depth untouched. We are getting much better at starting relationships and no better at holding them.
02 Three waves — and a longer memory
Networking has been redefined three times in twenty-five years:
Social networks turned the network into a visible, scalable, quantified asset. A contact became a number.
Covid and video dissolved geography and normalized meeting people you will never physically meet.
AI now mediates the interaction itself: what I write to you, how I approach you, and increasingly what I appear to know.
It is worth remembering that a century ago, networking was inseparable from place and body: guilds, chambers of commerce, markets, cafés, the parish. Your network was your territory. In the 1980s and 90s there was still no computer between two people who wanted to do business.
Each wave has extended reach and reduced friction — and each has removed something from the encounter. The question for this group is not whether this is good or bad, but what precisely was removed at each stage, and whether this last removal is different in kind.
03 What genuinely works — and I do not want this discussion to be only cautionary
AI functions as a social prosthesis. For someone shy, for a non-native speaker, for someone starting with no network at all — and when you start, you usually have nothing — AI lowers a barrier that used to be decisive. It helps you break the ice, phrase an approach, understand a foreign business culture before you walk into the room.
I experienced this directly while presenting in Japan and Taiwan: AI-assisted preparation and translation genuinely widened who could participate in rooms I would previously have been excluded from.
A network has historically been one of the most unequally distributed assets in any society. AI is, for the first time, redistributing access to it.
That deserves to be stated plainly before we discuss the risks.
04 Core hypothesis: AI generates the surface of a relationship, not its substance
This is where my domain converges with Theme B.
AI can now produce the appearance of competence: a flawless message, a confident summary of a field, an apparently informed opinion. But networking is the setting where knowledge without lived experience is exposed fastest. In a real conversation, expertise that has never been practiced collapses within minutes.
So the shift described in the summary for aesthetics (Horikawa, Souiki) applies equally to professional relations: as AI makes the surface of expertise free and infinite, value migrates to provenance — what you have actually built, whom you have actually served, who will vouch for you. In a world where anyone can sound like an expert, the scarce asset becomes verified experience and human endorsement.
A second point specific to my domain: AI is trained on the past. Networking is one of the primary sites where the genuinely new is produced. A market shift a founder felt last week, a failure nobody has published, a need not yet articulated. This is precisely the knowledge absent from every training corpus. Real encounter remains the mechanism by which new knowledge enters the world before it becomes data.
05 The gap: tools designed to connect can end up substituting
The risk I observe is enclosure: “I don't need people, I have my AI.” The tool built to help you approach others becomes comfortable enough that you stop approaching. This echoes the paradox noted in the summary for the auditory domain (Miyata) — greater accommodation of diversity coinciding with deeper isolation.
There is also a physical limit we should name plainly. You cannot shake a hand through a model. You cannot read a room, a hesitation, a silence. The social and bodily dimension of encounter is not a nice-to-have; in most cultures it is the mechanism by which trust is actually transferred.
This is not a fault of the technology. It is a design and cultural question we have not yet posed.
06 A practitioner's counter-model: Comindlab
I founded Comindlab as a new-generation coworking space built around collective intelligence — a place where entrepreneurs are less alone and less isolated, and where those who need to learn, train, or find financing can do so among others rather than in front of a screen. We host associations, training organizations, and companies seeking funding solutions.
What I observe there consistently: a group of humans in a room produces things that no individual and no model produces alone — because they see the reality in front of them, including everything that is not yet written down anywhere. I am an AI practitioner and a believer in the technology. I am also convinced that collective human intelligence remains, for now, the more powerful of the two.
07 Proposal for the Working Group
A simple evaluative criterion for AI in the relational domain, borrowing Tobias's distinction:
Does the tool end in a human encounter, or does it end in itself?
AI that increases the number and quality of real human meetings is a carrier of relational culture. AI that replaces the meeting substitutes for it. I would propose this as a design and policy test.
And a question I would like the group to consider: as professional identity and expertise become generatable, what would a verified-provenance layer for human relations look like — and who should govern it? This is the question I work on professionally, and I do not think it has a settled answer.
Submitted to the AI × Culture Working Group · Kyoto Culture Forum